Demonstration in Spinoza

Kant and Hegel both objected to Spinoza’s unusual presentation of his Ethics in something resembling the style of Euclid’s geometry. I think of philosophy mainly as interpretation rather than simple declaration, so I am broadly sympathetic to this point. On the other hand, I think Pierre Macherey is profoundly right when he emphasizes the non-foundationalist character of Spinoza’s thought.

The unique meaning Spinoza gives to “Substance” (not to be confused with its Aristotelian, Scholastic, Cartesian, or general early modern senses) is that of a complex relational whole that encompasses everything, rather than a separate starting point for deduction of the details of the world. Because of this, the apparent linearity of his development is just that — a mere appearance.

Hegel does not seem to recognize that Spinoza’s Substance resembles the relational whole of Force that Hegel himself developed in the Phenomenology. This is inseparable from an implicit notion of process in which relations of force are exhibited.

Macherey says Spinoza sees the world in terms of an infinite process, i.e., one without beginning or end or teleological structure (Hegel or Spinoza, p. 75).

(I would argue that neither Aristotle nor Hegel actually endows the world with teleological structure, though they each give ends a significance that Spinoza would deny. For Aristotle, it is particular beings in themselves that have ends. For Hegel, teleological development is a retrospectively meaningful interpretation, not an explanatory theory that could yield truth in advance. But for Spinoza, ends are either merely subjective, or involve an external providence that he explicitly rejects.)

It seems to me that the “point of view of eternity” that Spinoza associates with truth is actually intended to be appropriate to this infinite process. Spinoza points out that eternity does not properly mean a persistence in time that lasts forever, but rather a manner of subsistence that is entirely outside of — or independent of — the linear progression and falling away that characterizes time.

(Kant’s famous assertion of the “ideality of space and time”, which means that space and time are only necessary features of our empirical experience, is not inconsistent with Spinoza’s commendation of the point of view of eternity. Though it has other features Spinoza would be unlikely to accept, Kant’s “transcendental” as distinct from the empirical is thus to be viewed from a perspective not unlike Spinoza’s “point of view of eternity”.)

Spinoza wants to maintain that the order of causes and the order of reasons are the same. Whereas Aristotle deconstructs “cause” into a rich variety of kinds of “reasons why” (none of which resembles the early modern model of an impulse between billiard balls), Spinoza narrows the scope of “cause” to “efficient causes” in a sense that seems close to that of Suárez with inflections from Galilean physics, and suggests that true reasons are causes in this narrower sense. It seems to me that Spinoza’s “order of causes” resembles the infinite field of purely relational “force” that Hegel discusses in the Force and Understanding chapter.

Spinoza wants us to focus on efficient causes of things, but to do so mainly from the “point of view of eternity”. This takes us away from the event-oriented perspective of linear time, toward a consideration of general patterns of the interrelation of different kinds of means by which things end up as they concretely tend to do. In speaking of means rather than forces, I am tacitly substituting what I think is a properly Aristotelian notion of “efficient” cause for the meaning it historically seems to have had for Spinoza.

In pursuit of this, he takes up a stance toward demonstration that is actually like the one I see in Aristotle, in that it is more about improvement of our understanding through its practical exercise in inference than about proof of some truth assumed to be already understood (see also Demonstrative “Science”?). As Macherey puts it, for Spinoza “knowledge is not simply the unfolding of some established truth but the effective genesis of an understanding that nowhere precedes its realization” (p. 50). (Unlike Macherey, though, I think this is true for Aristotle and Hegel as well.)

Demonstration in both Aristotle’s and Spinoza’s sense is intended to improve our normative understanding of concepts by “showing” their inferential uses and points of application. It is only through their inferential use in the demonstrations that Spinoza’s nominal definitions and axioms acquire a meaning Spinoza would call “adequate”.

Place of a Preface

The preface to Hegel’s Phenomenology famously maintains that the conventional notion of a preface (e.g., “say what you are going to say”) is inapplicable to serious work in philosophy. There is an air of paradox about this, because in some sense he goes on to do what he just said was impossible. Hegel’s preface does summarize key conclusions of the book, but Hegel wants to make it clear that any such summary can at best be what Kant would call a dogmatic anticipation of real philosophical work yet to be done (in Brandom’s phrase, a “promissory note”).

I would note that this also reflects Hegel’s deep Aristotelianism. The way ideas are developed counts for much more than the way in which they are introduced. Aristotle did not follow the “say what you are going to say” model. Instead, he would begin with broad orienting remarks, a preliminary demarcation of subject matter, and a survey of common or leading views on the subject. Beginnings are the least certain part of a work; real substance emerges — if at all — from extensive development.

Of course it is possible to refer to an extensive development without producing or reproducing it in-line, but the relative soundness of such references depends on the soundness of the development and its applicability.

Simplicity is a pedagogical virtue that helps us on the uptake, but ultimately it cannot be the criterion of clarity. Real clarity comes from manifest interlinkages in a development that can be assessed independent of asserted conclusions (see Aristotelian Demonstration).

In his commentary, Harris says “The most important part here is Hegel’s insistence that the results of a science — whether it be philosophical or empirical — cannot be separated from the process (the Ausführung, or ‘execution’) by which they are reached” (Hegel’s Ladder I, p. 36). One of Hegel’s minor headings in the preface reads: “The principle is not the conclusion — against Formalism” (quoted on p. 48).

“We have a paradox here. The philosophical truth is absolute; but we have to hear it from one who is like ourselves. In this sphere, all particular situations are equally contingent. The philosopher, addressing her peers, will begin with this problem of how someone who accepts the finite human status can claim to say what is absolutely true — because that was precisely the problem that philosophy faced when the Phenomenology of Spirit was conceived. The philosopher and her peers, however, are not ‘in the midst of things’ in the way that the rest of us are. When she writes a book, she must take account of how we, the literate audience, are in medias res [in the midst of things]. Yet neither her situation nor ours is of any concern to philosophy as a systematic Science. For philosophy itself it is only the pure structure of ‘being in the midst of things’ that can be a possible starting point” (pp. 33-34).

Searching for a Middle Term

“But nothing, I think, prevents one from in a sense understanding and in a sense being ignorant of what one is learning” (Aristotle, Posterior Analytics; Complete Works revised Oxford edition vol. 1, p. 115). The kind of understanding spoken of here involves awareness “both that the explanation because of which the object is is its explanation, and that it is not possible for this to be otherwise” (ibid). To speak of the “explanation because of which” something is suggests that the concern is with states of affairs being some way, and the “not… otherwise” language further confirms this.

Following this is the famous criterion that demonstrative understanding depends on “things that are true and primitive and immediate and more familiar than and prior to and explanatory of the conclusion…. [T]here will be deduction even without these conditions, but there will not be demonstration, for it will not produce understanding” (ibid). The “more familiar than” part has sometimes been mistranslated as “better known than”, confusing what Aristotle carefully distinguishes as gnosis (personal acquaintance) and episteme (knowledge in a strong sense). I think this phrase is the key to the whole larger clause, giving it a pragmatic rather than foundationalist meaning. (Foundationalist claims only emerged later, with the Stoics and Descartes.) The pedagogical aim of demonstration is to use things that are more familiar to us — which for practical purposes we take to be true and primitive and immediate and prior and explanatory — to showcase reasons for things that are slightly less obvious.

Independent of these criteria for demonstration, the whole point of the syllogistic form is that the conclusion very “obviously” and necessarily follows, by a simple operation of composition on the premises (A => B and B => C, so A=> C). Once we have accepted both premises of a syllogism, the conclusion is already implicit, and that in an especially clear way. We will not reach any novel or unexpected conclusions by syllogism. It is a kind of canonical minimal inferential step, intended not to be profound but to be as simple and clear as possible.

(Contemporary category theory grounds all of mathematics on the notion of composable abstract dependencies, expressing complex dependencies as compositions of simpler ones. Its power depends on the fact that under a few carefully specified conditions expressing the properties of good composition, the composition of higher-order functions with internal conditional logic — and other even more general constructions — works in exactly the same way as composition of simple predications like “A is B“.)

Since a syllogism is designed to be a minimal inferential step, there is never a question of “searching” for the right conclusion. Rather, Aristotle speaks of searching for a “middle term” before an appropriate pair of premises is identified for syllogistic use. A middle term like B in the example above is the key ingredient in a syllogism, appearing both in the syntactically dependent position in one premise, and in the syntactically depended-upon position in the other premise, thus allowing the two to be composed together. This is a very simple example of mediation. Existence of a middle term B is what makes composition of the premises possible, and is therefore what makes pairings of premises appropriate for syllogistic use.

In many contexts, searching for a middle term can be understood as inventing an appropriate intermediate abstraction from available materials. If an existing abstraction is too broad to fit the case, we can add specifications until it does, and then optionally give the result a new name. All Aristotelian terms essentially are implied specifications; the names are just for convenience. Aristotle sometimes uses pure specifications as “nameless terms”.

Named abstractions function as shorthand for the potential inferences that they embody, enabling simple common-sense reasoning in ordinary language. We can become more clear about our thinking by using dialectic to unpack the implications of the abstractions embodied in our use of words. (See also Free Play; Practical Judgment.)

Demonstrative “Science”?

The “historiographical” notes on the history of philosophy I offer here from time to time are a sort of compromise. For much of my life, I’ve been very concerned with the fine grain of such history, and with casting a broad net encompassing many historical figures. Here, I made a strategic decision to focus instead on a mere handful of philosophers I consider most important.

Discussion of actualization in Hegel led to actualization in Aristotle, which led me to indulge my fascination with the Aristotelian commentary tradition. To the extent that it is possible to generalize about the historic readings discussed in the Greek, Arabic, Hebrew, and Latin commentaries, my own view of Aristotle is quite different on a number of key points, having more in common with some modern readings. Nonetheless, I am enormously impressed by the levels of sophistication shown by very many writers in this tradition.

I just mentioned al-Farabi again. As previously noted, al-Farabi (10th century CE) played a great historic role in the formulation of Arabic (and consequently, Hebrew and Latin) views of Aristotle. The Syrian Christians who did the majority of the translating of Aristotle to Arabic from Syriac had access to most of Aristotle’s works, but publicly only taught from the logical treatises. It was al-Farabi who initiated public teaching of the full range of Aristotelian philosophy in the Islamic world. He flourished during the so-called Islamic golden age, a time of tremendous interest in ancient learning not only by aristocrats but by many literate skilled crafts people. The political climate of the Islamic world at the time was much more embracing of secular learning than it came to be between the 13th and 19th centuries CE.

One unfortunate aspect of al-Farabi’s reading was a very strong privileging of a notion of demonstrative “science” over Aristotle’s own predominant use of dialectic in philosophical development. This was based on a reading of Aristotle’s Posterior Analytics as propounding a model of “science” as a deductive enterprise expected to result in certain knowledge, which is still dominant today, but which I (following a number of modern interpreters) think involves a misreading of the basic aims of Aristotelian demonstration.

The idea that Aristotle was fundamentally concerned to develop “sciences” yielding certain knowledge gave a more dogmatic cast to his whole work, which has been a contributing factor in common negative stereotypes of Aristotle. Many modern commentators who still accept this reading of Posterior Analytics have been puzzled by the huge gap between this and Aristotle’s actual practice throughout his works, which in fact is mainly dialectical. I think a careful reading of the Topics (on dialectic) and Posterior Analytics (on demonstration) with consultation of the Greek text on the originals of some key phrases yields a view that is far more consistent with Aristotle’s actual practice.

Demonstration is a pedagogical way of showing very clear reasons for certain kinds of conclusions. It works by assuming some premises are true, whereas dialectic makes no such assumption. Thus the only necessity that results from demonstration is the “hypothetical” one that if the premises are true, then the conclusion is also true. But the more important point in regard to the classic syllogistic form is that the common “middle term” that allows the major and minor premises to be both formally and materially composed together illuminates why we ought to consider it appropriate to assume the conclusion is true if we believe the premises are true.

Dialectic, as I have said, is cumulative, exploratory discursive reasoning about meanings in the absence of initial certainty. This is how Aristotle mainly approaches things. Dialectic implicitly relies on the same logical form of syllogistic argument explicitly used in demonstration, but Aristotle distinguishes dialectic and demonstration by whether premises are treated as hypotheses to be evaluated, or as hypothetically assumed “truths” to be interpreted.

It is also important to note that in the Latin scholastic tradition, the dogmatic trend resulting from wide acceptance of claims about demonstrative science was significantly mitigated by a strong counter-trend of evenhandedly analyzing arguments pro and con, which effectively revived a form of dialectic. (See also Foundations?; Fortunes of Aristotle; Scholastic Dialectic.)

Aristotelian Demonstration

Demonstration is literally a showing. For Aristotle, its main purpose is associated with learning and teaching, rather than proof. Its real objective is not Stoic or Cartesian certainty “that” something is true, but the clearest possible understanding of the substantive basis for definite conclusions, based on a grasping of reasons.

Aristotle’s main text dealing with demonstration, the Posterior Analytics, is not about epistemology or foundations of knowledge, although it touches on these topics. Rather, it is about the pragmatics of improving our informal semantic understanding by formal means.

For Aristotle, demonstration uses the same logical forms as dialectic, but unlike dialectic — which does not make assumptions ahead of time whether the hypotheses or opinions it examines are true, but focuses on explicating their inferential meaning — demonstration is about showing reasons and reasoning behind definite conclusions. Dialectic is a kind of conditional forward-looking interpretation based on consequences, while demonstration is a kind of backward-looking interpretation based on premises. Because demonstration’s practical purpose has to do with exhibiting the basis for definite conclusions, it necessarily seeks sound premises, or treats its premises as sound, whereas dialectic is indifferent to the soundness of the premises it analyzes in terms of their consequences.

We are said to know something in Aristotle’s stronger sense when we can clearly explain why it is the case, so demonstration is connected with knowledge. This connection has historically led to much misunderstanding. In the Arabic and Latin commentary traditions, demonstration was interpreted as proof. The Posterior Analytics was redeployed as an epistemological model for “science” based on formal deduction, understood as the paradigm for knowledge, while the role of dialectic and practical judgment in Aristotle was greatly downplayed. (See also Demonstrative “Science”?; Searching for a Middle Term; Plato and Aristotle Were Inferentialists; The Epistemic Modesty of Plato and Aristotle; Belief; Foundations?; Brandom on Truth.)

Kantian Discipline

The Discipline of Pure Reason chapter in Kant’s Critique of Pure Reason makes a number of important points, using the relation between reason and intuition introduced in the Transcendental Analytic. It ends up effectively advocating a form of discursive reasoning as essential to a Critical approach.

If we take a simple empirical concept like gold, no amount of analysis will tell us anything new about it, but he says we can take the matter of the corresponding perceptual intuition and initiate new perceptions of it that may tell us something new.

If we take a mathematical concept like a triangle, we can use it to rigorously construct an object in pure intuition, so that the object is nothing but our construction, with no other aspect.

However, he says, if we take a “transcendental” concept of a reality, substance, force, etc., it refers neither to an empirical nor to a pure intuition, but rather to a synthesis of empirical intuitions that is not itself an empirical intuition, and cannot be used to generate a pure intuition. This is related to Kant’s rejection of “intellectual” intuition. We are constantly tempted to act as if our preconscious syntheses of such abstractions referred to objects in the way that empirical and mathematical concepts do, each in their own way, but according to Kant’s analysis, they do not, because they are neither perceptual nor rigorously constructive.

All questions of what are in effect higher-order expressive classifications of syntheses of empirical intuitions belong to “rational cognition from concepts, which is called philosophical” (Cambridge edition, p.636, emphasis in original). This is again related to his rejection of the apparent simplicity and actual arbitrariness of intellectual intuition and its analogues like supposedly self-evident truth. It opens into the territory I have been calling semantic, and associating with a work of open-ended interpretation. (See also Discursive; Copernican; Dogmatism and Strife; Things In Themselves.)

I am more optimistic than Kant that something valuable — indeed priceless — can come from this sort of open-ended work of interpretation. Its open-endedness means no achieved result is ever beyond question, but I think we implicitly engage in this sort of “philosophical” interpretation every day of our lives, and have no choice in the matter. I also think serious ethical deliberation necessarily makes use of such interpretation, and again we have no choice in the matter. So, pragmatically speaking, defeasible interpretation is indispensable.

Kant goes on to polemicize against attempts to import a mathematical style of reasoning into philosophy, like Spinoza tried to do. Spinoza’s large-scale experiment with this in the Ethics I find fascinating, but ultimately artificial. It does make the inferential structure of his argument more explicit, and Pierre Macherey used this to great advantage in his five-volume French commentary on the Ethics. But there is a big difference between a pure mathematical construction — which can be interpreted without remainder by something like formal structural-operational semantics in the theory of programming languages, and so requires no defeasible interpretation of the sort mentioned above, on the one hand — and work involving concepts that can only be fully explicated by that sort of interpretation, on the other. Big parts of life — and all philosophy — are of the latter sort. So it seems Kant is ultimately right on this.

Kant points out that definition only has precise meaning in mathematics, and prefers to use a different word in other contexts. I make similar well-intentioned but admittedly opinionated recommendations about vocabulary, but what is most important is the conceptual difference. As long as we are clear about that, we can use the same word in more than one sense. As Aristotle would remind us, multiple senses of words are an inescapable feature of natural language.

Kant says that unlike the case of mathematics, in philosophy we should not put definitions first, except perhaps as a mere experiment. Again, he probably has Spinoza in mind, and again — personal fondness for Spinoza notwithstanding — I have to agree. (Macherey in his reading of Spinoza actually often goes in the reverse direction, interpreting the meaning of each part in terms of what it is used to “prove”, but the order of Spinoza’s own presentation most obviously suggests the kind of thing to which Kant is properly objecting.) More than anything else, meanings are what we seek in philosophical inquiry, so they cannot be just given at the start. We can certainly discuss or dialectically analyze stipulated meanings, but that is strictly secondary and subordinate to a larger interpretive work.

Following conventional practice, Kant allows for axioms in mathematics, but says they have no place in philosophy. He has in mind the older notion of axioms as supposedly self-evident truths. Contemporary mathematics has vastly multiplied alternative systems, and effectively treats axioms like stipulative definitions instead. If we have in mind axioms as self-evident truths, Kant’s point holds. If we have in mind axioms as stipulative definitions, then his point about stipulative definitions in philosophy applies to axioms as well.

A similar pattern holds for demonstration or proof. Mathematics for Kant always has to do with strict constructions, which do not apply in philosophy, where there is always matter for interpretation. (From the later 19th century, mathematicians began increasingly to invent theories that seemed to require nonconstructive assumptions — transfinite numbers, standard set theories, and so on. This is currently in flux again. Contrary to what was thought at an earlier time, it now appears that all valid “classical” mathematics, including transfinite numbers, can be expressed in a higher-order constructive formalism. Arguments are still raging about which style is better, but I am sympathetic to the constructive side.) Philosophical arguments are informally reasoned interpretations, not proofs.

Kant says that speculative thought in general, because it does not abide by these guidelines, unfortunately ends up full of what he does not hesitate to call dishonesty and hypocrisy. (When I occasionally ascribe honesty or dishonesty to a philosopher, it is with similar criteria in mind — especially the presence or absence of frank identification of speculation as such when it occurs. See also Likely Stories.)

The kind of philosophy I am recommending is concerned with explication of meanings, not a supposed generation of truths, so it is not speculative in Kant’s sense. What may not be obvious is just how large and vital the field of this sort of interpretation really is in life. The most common and compact form by which such interpretations are expressed in the small looks syntactically like ordinary assertion, and in ordinary social interaction, mistaking one for the other has little effect on communication. When the focus is not on practical communication but on improving our understanding, we have to step back and look at the larger context, in order to tell what is a speculative assertion and what is an interpretation expressed in the form of assertion. (See also Pure Reason, Metaphysics?; Three Logical Moments.)

(In the present endeavor, the great majority of what look like simple assertions are actually compact expressions of interpretations!)